The present invention relates to memory devices. More particularly, the invention relates to nonvolatile memory devices and read methods providing immunity from read disturbances.
Error detection and/or correction (ECC) schemes allow the recovery and restoration of data including one or more defective data bits. Errors are introduced into data via a number of different mechanisms during transfer, storage, and retrieval.
Many different ECC schemes have been used in variety of data storage and communication systems. Well-known ECC techniques include various coding schemes such as Reed-Solomon (RS), Hamming, Bose-Chaudhur-Hocquenghem (BCH), cyclic redundancy (CRC), and so forth. Using one or more of these ECC techniques, much defective data may be successfully identified and repaired.
In many contemporary data processing and storage systems, so-called “payload data”—the data having the desired information component—is stored in nonvolatile memory along with other data dedicated to the error detection and/or correction. This type of collateral data is commonly referred to as ECC data and is derived from the payload data. Once stored with the payload data, ECC data may be used to correct errant data bits generated during a read operation directed to the nonvolatile memory. The nature and quantity of the ECC data defines the actual number of errant data bits that may be corrected. By employing a competent ECC technique and coincidentally storing ECC data, many bit errors associated with read operations may be quickly and efficiently repaired without the need to use more onerous remedies such as block replacement.